Revision of Pilargis de Saint‐Joseph, 1899 (Annelida, Polychaeta, Pilargidae)

Pilargis de Saint‐Joseph, 1899 includes seven species and one subspecies: P. berkeleyae Monro, 1933, P. maculata Hartman, 1947, P. modesta Intes and le Loeuff, 1975, P. mohri Gallardo, 1968, P. papillata Rasmussen, 1973, P. tardigrada (Webster, 1879), P. verrucosa de Saint‐Joseph, 1899, and P. verrucosa pacifica Uschakov, 1955. Two species (P. verrucosa and P. berkeleyae) have been recorded from widespread localities. However, no comparison with type material has been done, and there might be some different forms under the same name. We have reviewed all the available type material in order to clarify the taxonomy of this group. A critical analysis of morphological features and a standardization of their structure and variability have been attempted. Six species are redescribed, one species is characterized after published accounts, and one subspecies is elevated to species. Four are described as new: Pilargis angeli n. sp., P. cholae n. sp., P. rozbaczyloi n. sp., and P. wolfi n. sp. Three others based on damaged material are briefly characterized. A key for Pilargis species is included.


Introduction
Pilargis de Saint-Joseph, 1899 with Pilargis verrucosa de Saint-Joseph, 1899, as the type species, was proposed in two different publications and in both cases a new family (Pilargidae) was established. The short publication (de Saint-Joseph 1899a) was the proposal of the family, the inclusion of Phronia tardigrada Webster, 1879, within it, as well as the definition of a new genus, Pilargis. The paper has a brief reference to the type species, P. verrucosa, and mentions the later publication. The second publication came a few months later and was a faunistic report on the polychaetes from Brest and Paimpol (de Saint-Joseph 1899b); besides some minor punctuation or redaction differences, family and genus diagnosis are identical. In the second paper, the species description is complete and illustrated, and the corresponding section in the paper finishes with a short explanation on the reasons to place both genera in a distinct family. However, the illustration does not give enough detail of the anterior end, and there was some confusion on the naming of its may change by pharynx eversion; palpostyles are always small, sometimes slightly larger than surrounding verrucae, and can be placed towards the inner margin or by the center of the palp. Palps are biarticulate, although palpostyles may be tiny and difficult to detect among large verrucae, or eroded during sample treatment, so a careful ventral or frontal examination is often required to detect them.

Lateral antennae
There are two simple lateral antennae, usually shorter than palps. They may arise from lateral ridges or can be placed at about the same level as the middle of the prostomium, extending along different proportions of palp length. This insertion is easily noticed by observing the anterior reach of the antennae, so they can be anterior if they reach the palp tip ( Figure 11A), or median if they remain shorter than palps ( Figure 7A). The relative length and position become fixed early in ontogeny since juveniles (see P. verrucosa below) show a similar pattern in position and size proportions.

Tentacular segment
The tentacular segment carries two pairs of tentacular cirri. It is anteriorly reduced, leaving some room for a mid-dorsal nuchal area. Fauvel (1920) regarded it as a brain lobe, but it has not been detailed afterwards; in his drawings, its position and elevation resemble the sensory lobe that is present in other polychaete families (e.g. Amphinomidae), but this could be an artifact of contraction. The tentacular cirri are placed slightly differently; the dorsal cirri are placed slightly ahead of the ventral ones so, even though they are of about the same size, the dorsal may look longer.

First setiger
There may be some differences in the size relationship of the first dorsal cirri, in relation to tentacular cirri, and to dorsal cirri in following setigers. Thus, they are always larger than tentacular cirri and can be larger, about the same size, or smaller than those present in following setigers.

Parapodia
Parapodia are swollen, massive lobes with little muscular development. They are hollow and enteric caeca penetrate deeply into them ( Figure 8F). Parapodia have dorsal cirri, neurosetal lobes with setae, and ventral cirri. Dorsal cirri can be separated in most species very easily into cirrophore and cirrostyles; the size relationships between them can be useful, but since their proportions may change throughout the body, anterior and posterior parapodia must be observed. The cirrophores can be massive, much wider than the cirrostyles ( Figure 5C, D), glandular and colored, while the cirrostyles can be as wide as the cirrophore (Figures 1C, 14E) or much narrower. The parapodial lobe may show some folding along its external wall; the pattern of folds, either transversal or longitudinal, may indicate the degree of wall thickness or amount of muscle layers in it. The ventral cirri can extend beyond the setal lobe (without setae), no matter how contracted the setal lobe is, or it can be shorter than it. However, since the setal lobe is employed for walking, it has a large amount of muscles, and its appearance can vary depending on its contraction degree. In comparison, parapodial cirri are not so contractile and its shape is more conservative and useful for taxonomic purposes.

Parapodial glands
Glands are present in dorsal cirri, and sometimes over the ventral surface of the parapodia and body wall as well. Their arrangement can be concentrated on the dorsal (often anterior) surface of the cirrophore ( Figure 8C, D), or they can be placed more deeply into the dermis. Further, there might be smaller glands on the ventral surface of the parapodial lobe ( Figure 13C, D). These glands have spherical or polyhedrical gland cells, often dark colored. Sometimes, pigment may fade but the structure can be detected in the compound microscope, and could be seen more easily using methyl-green staining (Winsnes 1985). The gland area is rather small in juveniles but becomes larger through development; however, its relative position is conservative, making it a useful character to separate species. In the dorsal cirrostyle, the glands can be rounded or club-shaped, much less pigmented than other glands.

Neurosetae
The brittle setae can be capillaries or limbates. The former are cylindrical, smooth and can be long and distally curved, while the latter are basally cylindrical and distally limbate, straight or curved, sometimes twisted over itself, looking like a pseudocompound setae ( Figure 10C). The distal region is not cylindrical; it is a triangular prism, and the cutting edge can be finely spinulose or clearly denticulated ( Figure 18E). Some species have distally curved neurosetae, sometimes with a larger distal denticle, making the setae appear bidentate. These details have been illustrated for some species since Hartman (1947), and have been recently studied by scanning electron microscopy (Cañ ete et al. 1990), indicating that the main tooth is not spoon-shaped, and the distal denticle is not shovelshaped, as has been shown before (Hartman 1947). The relative exposure of setae should be taken into account; if the distal prismatic portion is the only one exposed, then the setae could appear short, completely limbate, and either spinulose or denticulate. These features would be less eroded and more easily seen in shorter setae; longer setae would appear medially limbate, and in larger organisms, setae would look subdistally limbate when fully exposed. Further, setal tips erode or adsorb foreign materials very easily, and their fine denticles or limbus spines are fragile, and they can be easily eroded or covered by adsorpted materials too, making its detection more difficult. Therefore, neurosetae should be called either straight or curved limbate capillaries; however, because they are very brittle and easily adsorb foreign materials, their use should be de-emphasized in separating species in the genus. The furcate setae reported found by Fauvel (1920Fauvel ( , repeated: 1923Fauvel ( , 1934Fauvel ( , 1936, were not seen by Katzmann et al. (1974), nor by us. They are not present in the genus.

Posterior end
Although members of Pilargis fragment easily, the posterior end may provide some useful taxonomic characters. The number of prepygidial asetigers would be of minor importance. The pygidium may be swollen as an anal bulb, can be as wide as the preceding body section, or could be smoothly tapered, and these character states are shown in the same species, depending on the development or state of regeneration. Anal cirri can be present or absent, although sometimes some large verrucae could be confused with short anal cirri. To be regarded as true anal cirri, they should be at least twice as long as surrounding verrucae. However, in some specimens of the same species, some had pygidium with cirri while others lacked them ( Figure 12B-D); thus, we should de-emphasize their use as a distinguishing feature.

Brain lobes
The brain shape of pilargids was employed by Fitzhugh and Wolf (1990). They found that it has a pair of large posterior lobes, often darkly colored, that can project over setigers 1-2. They regarded a pair of lateral dark glands as nuchal organs and one pair of depressions in the postectal corners of the prostomium were referred to as nuchal slits. As will be shown below, there could be other dark internal masses, especially in the tentacular segment. Since they did not find any cilia in nuchal organs or slits, their proposal to name these structures should be regarded as tentative. There can be differences in the size and pigmentation degree of the brain posterior lobes and they could be used to separate similar species, though pigment fades off rather easily. The posterior lobes can be globose or fusiform, and can be parallel to the main body axis or have an oblique orientation, and they can be colorless (
The proposal of the subfamilies Synelminae and Sigambrinae was made by comparative morphology methods (Salazar-Vallejo 1987). However, it failed to follow the Principle of Coordination (International Commission on Zoological Nomenclature (ICZN) 1999, Art. 36), thus it was consequently modified by using Pilarginae instead of Sigambrinae by Salazar-Vallejo and Orensanz (1991). After the subfamilies proposal by Salazar-Vallejo (1987), there have been two independent evaluations of the intergeneric cladistic affinities in the Pilargidae. The first was by Fitzhugh and Wolf (1990), and the second by Licher and Westheide (1994), and despite their conclusions, the subfamilies are apparently well delineated in their cladograms. Fitzhugh and Wolf (1990, p 16, Appendix 2) regarded three genera as lacking notopodial spines: Loandalia Monro, 1936, Parandalia Emerson andFauchald, 1971, and Pilargis. However, as has been shown in several publications (Salazar-Vallejo 1987;Salazar-Vallejo and Orensanz 1991;Blake 1994), the former two do have notopodial spines, some species even have a few thin capillary setae along with them. Further, these two genera have been shown to be synonyms (Salazar-Vallejo 1998), so among the genera included by Fitzhugh and Wolf (1990), only Pilargis completely lacks notospines or notohooks. Further, Synelmis (5Glyphohesione) klatti was regarded as a member of the family (but it belongs elsewhere), and if we understand the proposed subfamilies as represented by genera 1-6 for Pilarginae against 7-12 for Synelminae, they were grouped in two out of three trees (Fitzhugh and Wolf 1990, Figures 4, 5). Licher and Westheide (1994, p 228, Table 1, p 232, Figure 4) indicated that two genera lack notopodia (although they probably meant notosetae): Pilargis and Otopsis Ditlevsen, 1917, and they differ especially because Otopsis has a smooth integument, median antenna, and its palps are regarded as simple, while in Pilargis integument is verrucose, median antenna is missing, and palps are biarticulated. Further, in their single cladogram, they found that Glyphohesione was basal and the two subfamilies are better delineated, since they found a single tree, with the sole exception of Otopsis, which was wrongly coded (see Licher and Westheide (1994, Figure 4, 19d) because they indicated lack of notopodia, but Otopsis only lacks notosetae, and has large dorsal cirri. These features bring it in close proximity to, and may eventually fall within Pilargis. Anyway, after Jenner (2002Jenner ( , 2004, there is a widespread problem in character definition and character coding, so we are still far from a robust conclusion or rejection of the subfamily groupings. Regarding the differences between Pilargis and Otopsis, as will be shown below, there is a marked variation in the verrucal density or relative size, and in some species they are markedly reduced in size or body coverage (see below). The number of antennae has been rejected as a generic diagnostic feature by Pettibone (1966, p 164) for separating Ancistargis Jones, 1961 from Ancistrosyllis McIntosh, 1879 (a contrasting view was briefly exposed by Emerson andFauchald 1971, andby Licher andWestheide 1994, p 233) for Litocorsa Pearson, 1970. However, the median antenna in Otopsis is very large, at least in the type species (O. longipes Ditlevsen, 1917). Further, palpostyles may be very small and difficult to find, or they can be eroded, making any distinction of simple against biarticulate rather difficult to make. However, another species (O. kurilensis Uschakov, 1971) has a tiny median antenna and biarticulated palps, which indicates a much closer similarity with Pilargis. If we follow the same approach of disregarding the number of antennae as a distinguishing feature, then Otopsis might have to be included in Pilargis since no other feature separates them, as was indicated by Hartman (1947, p 483, footnote 1). This is an interesting question that will be addressed in a forthcoming contribution.

Redescription
Since the original description and Fauvel's redescription are good, a few comments will be added on the type material.
Female syntype (A278fA). Three fragments, pale, anterior one 12 mm long, 3.5 mm wide, 39 setigers; median large portion 10 mm long, 3.2 mm wide, 27 setigers; short median portion 3 mm long, 3 mm wide, four setigers. First dorsal cirri larger than the one of setiger 2 ( Figure 1A); dorsal cirri much larger than ventral one, 1.5-2.0 times longer and 2-5 times wider, can be separated in cirrophore and cirrostyles, about equal-sized but cirrophore much wider ( Figure 1C, D). No posterior end available; larger verrucae conical or mammiliform, those present on dorsal cirrostyles, ventral cirri, and ventral surface low cylindrical; once cleaned, none has any filament through the transparent integument, as was depicted in the original Figure 17. Bifid setae could not be confirmed because of the amount of adsorpted salt over them; this has already been noticed by Fauvel (1925, p 90).
Male syntype (A278m). Two fragments, pale, anterior one 44 mm long, 4 mm wide, 124 setigers; median portion 18 mm long, 4 mm wide, 18 setigers. Both partly dehydrated, show a deep ventral furrow, stiff; anterior fragment previously dissected ventro-longitudinally, from slightly after the mouth to setiger 42, but blade went through both body walls, several parapodia removed. Some neurosetae free of adsorbed materials, smooth limbate capillaries distally curved, a main and an accessory tooth. The difference that de Saint-Joseph noticed between male and female was probably due to the different degree of dehydration, which changes the general body appearance and the relative proportions of the parapodia. Larger verrucae over middorsal and parapodial base areas, though the whole dorsum is covered by verrucae. Sperm abundant in coelom, head globose, about 3 mm long.
Juveniles. Complete specimen (ECOSUR), 9.5 mm long, 0.4 mm wide, without any median expansion, 52 setigers, four asetigers. Integument with sparse verrucae, except in the anterior end. Body with a series of glandular masses dorsally and ventrally in parapodial bases, in setigers 3-35 about the same size, pigmentation fading towards setiger 30,  disappear in setiger 35, reappear only in setiger 43; they disappear in the posterior region, although there is a dark lateral gland in the pygidial bulb.
Prostomium with biarticulated palps, palpostyles central directed forward; lateral antennae inserted on the base of the palps, slightly pass prostomial margin but not palp tips ( Figure 2A). Tentacular segment projected forwards; tentacular cirri with verrucae, dorsal one slightly larger than ventral one. First setiger with dorsal cirri larger than dorsal tentacular cirri and dorsal cirri on setiger 2. Ventral cirri longer than setal lobe. Parapodia with acuminate cirri, dorsal cirri larger than ventral one, both larger than setal lobe.
Posterior end with four asetigers, pygidial bulb with abundant verrucae, from its medial portion towards the posterior margin, two ventrolateral cirri, over 10 times the size of adjacent verrucae ( Figure 2B). Most verrucae over anterior end and over parapodial cirri are not flat; rather bifid or trifid, tend to be regularly placed over the body; three ones over lateral margins of tentacular segment, two to three over dorsal cirrostyles and one subdistal in ventral cirri (Figure 2A, insert). Brain elongated, posteriorly bilobed, reaches about the middle of second setiger. Gut straight; lateral diverticula not seen.
Another juvenile (SMF) completely transparent ( Figure 2C, D); two large dorsal ovoid dark gland masses from setiger 3, and two other ones over the ventral surface, continue towards some five setigers before pygidium, giving the specimen a distinct spotted regular pattern. Tentacular cirri directed anterolaterally, first dorsal cirri about the same size as second setiger ones. Posterior end with two asetigers, short pygidial bulb, two lateral anal cirri, all with large verrucae. Gut diverticula from setiger 1.

Variability
The insertion of the lateral antennae, as well as size relationships between anterior end cirri, are consistent since they become fixed in small specimens. However, the abundance and shape of verrucae differ from larger specimens; dorsal verrucae shape differs, and those present over cirri tend to be smaller in larger specimens. A comparison of parapodial development in different organisms is shown in Figure 3; cirri are always more or less globose, blunt, and dorsal larger than ventral ones. Dorsal verrucae are always larger ( Figure 3B-F), inverted cone-or dune-shaped ( Figure 3E, insert), than those placed over the cirrostyles; these are often distally expanded ( Figure 3F, insert), and arranged in a single longitudinal line over the cirrostyles.
The dark glandular material is present just before the dorsal cirrophore ( Figure 2A, C) and disappears in larger specimens; first appearing as small globular structures and later they cannot be seen again. Likewise, body outline changes a lot since the juvenile is almost cylindrical or slightly tapered towards both ends, while the adult is markedly wider in the median region than towards any end. The posterior end is rather variable although the basic structure is more or less retained, the anal cirri might regenerate and the swelling of the pygidial bulb ( Figure 3A) may change with contraction. Gut diverticula are seen in every setiger in juveniles ( Figure 2C) and continue to the last prepygidial setiger ( Figure 2D) . Fauvel incorrectly stated that verrucae are restricted to the dorsal side of dorsal cirri, since they are present, though smaller, over the ventral parapodia surface. Anterior end lacks the caruncle that Fauvel had noticed; it has a rather smooth surface without large verrucae. Further, the posterior brain lobes are slightly darker, placed inside setiger 1, and somewhat fusiform pointing anterolaterally. Posterior end as a pygidial bulb heavily covered by large truncate verrucae, but may lack anal cirri; neurosetae masked by adsorpted materials but a few of them show an almost smooth limbus and a distally curved bifid tip, as was illustrated for it.

Discussion
Pilargis verrucosa has no sexual dimorphism, as has been already stated by Katzmann et al. (1974, p 21); however, it includes two different morphs in the Bay of Biscay, depending on the abundance and size of verrucae: the typical morph has abundant large verrucae over the back and parapodia, and the atypical morph has few and smaller verrucae. There are other differences in the anterior end and parapodial cirri; however, since sampling depth was not recorded, this variation remains unexplained. Therefore, those morphs with few verrucae are herein regarded as members of P. modesta Intes and le , which is a southern species; it seems that these latter authors did not include the paper by Katzmann et al. (1974) because it was in press.

Distribution
France, Iberian Peninsula and Mediterranean Sea, in 2-300 m depth.
First setiger with dorsal cirri longer than tentacular cirri, longer than following cirri. Anterior parapodia ( Figure 4B) with some verrucae over the back and cirrostyles; dorsal cirrophore thick, globose; no traces of large glandular areas in the middle of the cirrophore nor epidermal glands, but a diffuse spotted pattern of glands in parapodia of median fragments. Ventral cirri elongate, about one-third as long as dorsal cirri; dorsal cirrostyles cirriform, very long, some with a distal swelling, 1.2-1.5 times cirrophore length. Median and posterior parapodia similar ( Figure 4C-D), without verrucae; ventral cirri shorter, cirriform, longer than setal lobe.
Pilargis angeli n. sp. is very similar to P. tardigrada (Webster); they differ because the first have verrucae on its anterior end, while the second has verrucae restricted to parapodial lobes. Further, the size of cirrostyles is different, being very long in P. angeli n. sp., and conical and shorter in P. tardigrada.

Etymology
This species is named after our colleague J. Angel de Leó n-Gonzalez, who has been working a lot on Eastern Pacific polychaetes in general, and who provided us with the type material for this species.

Type locality
Off the Western coast of Baja California Sur, México.

Material examined
Eastern Pacific Ocean: one specimen (LACMNH), collected during the Southern California Bureau of Land Management Study, accession number 85401, code BF1, no further data.
Parapodia elongated. First dorsal cirri as long as second dorsal cirri. Dorsal cirrophore elongated truncate cone, no glandular material visible either on the surface or below the integument. No pigment spots over ventral parapodial lobe. Dorsal cirrostyles digitate, elongated, no distinction observable from cirrophore. Ventral cirri cirriform, elongated, longer than setal lobe, about as long as dorsal cirri. Neurosetae include very long capillaries and smaller finely spinulose bidentates.
Pygidium rounded, two conical, ventrolateral anal cirri, as long as last three to four asetigers. Pharynx conical, wider anteriorly, as long as first three setigers. Gut diverticula penetrate about half parapodial lobe in setigers 1-9; from setiger 10, diverticula not yet formed; gut empty but in posterior region with fine particles and sediments.

Discussion
By the presence of elongated capillary setae and by the incomplete development of the gut diverticula and posterior end, this form is an early juvenile or a just settled postlarva.
Whether it belongs to other species in the region like P. berkeleyae or P. maculata (see below) is difficult to say, because there is no glandular development. However, because of the long dorsal cirrophore, it may be conspecific with P. angeli n. sp., but the setae differ. There are no other records of juvenile or postlarval Pilargis, and the few records available are on Ancistrosyllis (Bhaud 1974;Blake 1975); it is interesting that this juvenile has antennae placed towards the posterior end of the prostomium; they could migrate anteriorly as the peristomium grows more laterally and fuses with the prostomium. Tentacular cirri and parapodial cirri are little specialized too, they undertake differential growth since in adults, dorsal tentacular cirri is longer than ventral one, dorsal cirri of first setiger becomes larger than those present in setiger 2, and ventral cirri is smaller than dorsal cirri. More and better materials would help in solving its affinity or specific placement. Monro, 1933 (
Prostomium fused with peristomium; palps globose, palpostyles tiny. Antennae cirriform, placed dorsally on the basis of palps. Peristomium 1.3 times longer than first setiger. Tentacular cirri cirriform, as wide as dorsal cirri of setiger 1. First setiger with dorsal cirri about twice as long as dorsal tentacular cirri ( Figure 5B). Dorsal cirri of setiger 1 three times as long as second dorsal cirri; right second dorsal cirri missing. All parapodia with dorsal cirrophore globose; anterior setigers with few small dark glands ( Figure 5C), concentrated under the epidermis. Median parapodia with cirrophore turgent, thick, with dark glandular area completely filling the cirrophore. Posterior parapodia with gland mass perimeter less heavily pigmented ( Figure 5D); glandular material concentrated slightly below the insertion of cirrostyles. No additional pigmented materials visible. Dorsal cirrostyles thick, digitate, about one-third to one-fourth as long as cirrophore. Ventral cirri thin, cirriform, can be as long as acicular lobe in anterior setigers; in most cases not surpassing the acicular lobe. Superior neurosetae mostly smooth capillaries, inferior bundle setae limbate, finely spinulose, distally bifid.
Pharynx everted, globose, transparent, broken in three sectors; it has two lobes clearly separated from the inside. Right setiger 8 has been previously removed; lumen can be seen from the outside; gut diverticula may start some setigers before it. Non-type mature female (LACNHM-1714) with eggs in median and posterior parapodia from setiger 240, one to three large eggs (200 mm each) per parapodium, others in coelom, spread out after parapodial removal.

Variation
Non-type specimen is a very large anterior fragment of a mature female, 170 mm long, 2 mm wide at setiger 20, 482 setigers. In spite of its large size, the glandular portion over anterior surface of dorsal cirrophore is small, and the integument is rough, with sparse small verrucae, concentrated over the anterior end and cirrostyles. Parapodia very corrugated with longitudinal and transverse streaks ( Figure 5E). Dorsal cirrophore well developed, not clearly cut from parapodial lobe, few spherical glands inside it, without pigmentation; cirrostyles digitate, 1.5 times longer than wide, with few verrucae concentrated over its dorsal surface. Ventral cirri directed laterally, digitate, twi to three times longer than wide, surpassing setal lobe. Neurosetae long smooth capillaries and limbates finely spinulose, bidentate. Everted pharynx broken, no internal septum was seen.
One juvenile with large globose structures in anterior segments ( Figure 6A), could be yolk granules and, if so, indicative of a lecitotrophic development. Besides pigmented glandular areas in dorsal cirrophore, many dark granules irregularly placed over the back ( Figure 6B), while ventral surface has black spots restricted to parapodial bases ( Figure 6C).

Discussion
The relative size and abundance of verrucae do not change with sexual maturity. Further, the development and position of the glands over the parapodia are very consistent and useful characters; thus the more closely allied Californian species, P. berkeleyae Monro, 1933 and P. maculata, cannot be separated using verrucae abundance but they can more easily be identified using the pattern of glandular development on parapodia. The former has glands that extend over the cirrophore, occupying from the subsurface to the inner portion of the cirrophore, while the latter has glands more or less restricted to the anterior surface of parapodial lobes, and they are rather superficial. The record by Imajima (1987, p 162) may not belong to this species because of the presence of abundant large verrucae (see Imajima 1987, Figure 7), and the insertion of those illustrations by Blake (1994) might promote confusion. Further, the record of a commensal worm in Chaetopterus tubes (Britayev 1993), implies a species different from that found by Imajima, judging by the size and abundance of verrucae (seen in his figures). They should be compared with P. pacifica Uschakov, 1955, originally described from Northern Japan Sea.

Distribution
Described from Friday Harbor, Washington, it has been documented as far south as Southern California. Other records are questionable because of the confusion regarding the development of dorsal verrucae, and glandular patterns in dorsal cirrophores. Pilargis cholae n. sp. (Figure 7) Pilargis berkeleyae Wolf 1984, p 29.26-29.28, Figures 29.23-29.24 (non Monro, 1933.

Description
Holotype anterior fragment, 25 mm long, 1.8 mm wide, 101 setigers. Verrucae very conspicuous, large capitate ones dorsally, smaller ones ventrally, the latter especially in the anterior region; posterior region almost smooth.
Prostomium completely fused to peristomium; palps biarticulated, damaged. Lateral antennae small, hardly seen because of abundant verrucae, observed in a ventral view of paratype; palps slightly damaged ( Figure 7A). Two pairs of tentacular cirri, both directed anteriorly, covered with small verrucae.
Parapodia with large fusiform dorsal cirri, abundant large verrucae over the back, smaller ones over the tiny cirrostyles. First setiger with dorsal cirri larger than the one of setiger 2. Anterior parapodia ( Figure 7C) with large fusiform dorsal cirri, 1.3 times longer than ventral cirri, about four times wider; ventral surface with rounded verrucae, about one-fifth as large as the largest dorsal verrucae. Median and posterior parapodia with larger dorsal cirri ( Figure 7D, E), about two times as long as ventral cirri, four to five times wider. Ventral surface with tiny verrucae or smooth. Neurosetae include smooth capillaries and limbates, finely spinulose, very slightly bent distally, bidentate.
Posterior end (seen in paratype) with a pygidial bulb, as long as last two asetigers, distally covered by verrucae, in an expansion that projects from a short smooth neck-like portion; two lateroventral anal cirri, right one duplicate, all covered by verrucae ( Figure 7B). Brain posterior lobes with heavy pigmentation, slightly pass first setiger. Other pigmented masses placed in bases of left parapodium 1, and tentacular cirri.

Discussion
Pilargis cholae n. sp. differs from P. berkeleyae by having abundant large verrucae dorsally, and by lacking any large glandular region in the cirrophore. It rather resembles P. verrucosa, but cirri development is very different, being massive in P. verrucosa and more elongate, especially the ventral cirri, in P. cholae n. sp. It differs from P. papillata by lacking 8-shaped verrucae, especially on the base of dorsal cirri. The description by Wolf (1984) is accurate, except that the anal cirri were indicated shorter than what they are.

Etymology
This species is named after M. Soledad Jiménez-Cueto, a former colleague who sampled and processed the material on which this description is mainly based. The specific name is formed after her nickname.

Type locality
Off Panama City, Western coast of Florida, Gulf of Mexico.

Distribution
Restricted to two localities in Western Florida, in soft bottoms in 37-53 m depth, and in shallow (0-1.5 m) seagrass (Thalassia testudinum) beds, Laguna Nichupte, Quintana Roo, México. Hartman, 1947 ( Figure  Hammond (148 with four fragments, two anterior ones; larger selected as lectotype; smaller anterior fragment 2 mm long, 0.9 mm wide in last setiger, 13 setigers; larger median fragment 44 mm long, 1 mm wide around its middle, ca. 195 setigers; some segments with a large egg in ventro-posterior space in the segment; a smaller fragment 8 mm long, 0.9 mm wide about its middle, 40 setigers. Paralectotype is a posterior and a median fragment, probably coming from different specimens, because they differ in their condition.
First setiger with dorsal cirri about as long as dorsal tentacular cirri, two times longer than dorsal cirri of second setiger. Parapodia with few contraction lines, short cirrophore and cirrostyles; dorsal cirrophore truncated. Pigmented dark glands below integument, arranged in an oblique elliptical area, concentrated over anterior surface of cirrophore ( Figure 8C), starting from setiger 1 or 2; first as few glands, in following setigers individual size enlarges and increase their number. Dorsal cirrostyles digitate, about half the length of cirrophore ( Figure 8D). Ventral cirri cirriform, smaller than acicular lobe. Ventral dark small glands, one or two per parapodium, are placed towards the parapodial base. Most setae broken; setae include superior large, inferior smaller, bidentate laterally spinulose capillaries. Sometimes in anterior setigers, long setae may look smooth but they are finely spinulose ( Figure 8E).
Posterior end (observed in a paralectotype) with pygidium rugose, without anal cirri. Pharynx not everted. Gut diverticula not seen by transparency, but present along the body ( Figure 8F). Some segments have few eggs inside them, each 90-140 mm. (1966) stated that this species is conspecific with P. berkeleyae Monro, but both can be recognized using glandular development over the dorsal cirrophore, being superficial and oval-shaped in this species, while in P. berkeleyae glands tend to fill the cirrophore and provide a dark color for it. The terms type and cotype should not be used in nomenclature (ICZN 1999, Recomm. 73E); further, the type cannot be regarded as the holotype because it consists of two fragments, which apparently are not coming from the same specimen. One specimen formerly included as part of the cotype is herein being designated as the lectotype; it fits the original description.

Distribution
Restricted to the California current coastal ecosystem.

Material examined
Northwestern Africa: holotype (MNHN-A856) of Pilargis modesta Intes and le Loeuff, 1975, coll. A. Intes and P. le Loeuff, off  Body with many small verrucae over anterior end, less abundant in median segments, almost disappear completely by posterior end. Verrucae smaller, less abundant over parapodial lobes in median and posterior segments. Prostomium dorsally free from prostomium; palps large biarticulated, directed ventrally. Lateral antennae over anterior prostomial margin; few tiny dark internal spherical structures. Tentacular cirri directed anteroventrally, both of about the same size ( Figure 9A).
First setiger with dorsal cirri larger than the one of setiger 2. Parapodia with acuminate, globose dorsal cirri, digitate ventral cirri. Anterior parapodia with large dorsal cirri, about 1.3 times as long as ventral cirri; cirrophore massive with two to three middorsal verrucae, few spherical internal glands, cirrostyles small, about one-fifth as long as cirrophore. Setal lobe rounded with projected acicular lobe ( Figure 9C). Median setigers with less abundant verrucae ( Figure 9D), often setal lobe deeply invaginated in parapodium; dorsal cirri become larger, 1.5 times longer than ventral ones; size proportions between cirrophore and cirrostyles remain constant posteriorly but glands in cirrophore become more abundant. Posterior parapodia with large ova, each ca. 150 mm ( Figure 9E), can be seen in holotype from setiger 87, continue to the end of fragment, more abundant by setiger 200 ( Figure 9B). Neurosetae long or short, sometimes curled; few long capillaries, most limbates very finely spinulose, limbus almost smooth.

Discussion
The inclusion of some records of P. verrucosa in P. modesta was briefly discussed under the latter species. The description of Pilargis modesta Intes and le  was overlooked by Kirkegaard (1983), who identified his material using the revision by Pettibone (1966). The name was not explained by Intes and le Loeuff; it might indicate the small size of animals, in comparison with P. berkeleyae, or more likely the weak development of dorsal verrucae. The original illustration shows very large divergent tentacular cirri but they are actually directed forwards and are rather short. There are two interesting features regarding the verrucae in P. modesta: they are linked to the internal layer through a very thin cuticle tunnel, without the internal development seen in other species, and fungiform verrucae are possibly artifacts, as they are not necessarily fixed over a cuticle tunnel. The verrucae in this species are very flat, with a tiny discoid outer layer and a hardly visible thin cuticle tunnel. Further, in two notes related to P. modesta (Intes and le Loeuff 1975, p 299;Kirkegaard 1983, p 211), neurosetae were described as spinulose, but all have smooth limbus.
Pilargis modesta differs from P. berkeleyae in that the former has conical or slightly swollen dorsal cirrophores (two to three times wider than the cirrostyles), and little glandular development, while in P. berkeleyae the dorsal cirrophore is globose (four to five times wider than the cirrostyles).

Discussion
It was identified as, and formerly characterized in relation to P. berkeleyae, but it does not belong in it. The main difference is that this specimen does not have a large cirrophore nor do the pigmented glands layer in it ( Figure 9F). Further, its lower shorter neurosetae are clearly spinulose and spines are large enough to be seen with low magnification (106), while in P. berkeleyae neurosetae are very slightly spinulose. This seems to be an undescribed species, close to P. modesta, but different from it by having spinulose neurosetae. More and better material should be gathered in order to propose a new name.

Redescription
Holotype includes a long anterior, and a posterior fragment; anterior one 96 mm long, 1.0 mm wide, 179 setigers; posterior one 14 mm long, 1.0 mm wide, 33 setigers (plus a few more in regeneration). Body pale, tapered towards anterior and posterior end, slightly wider towards the middle of the body; abundant verrucae. Anterior and posterior ends with verrucae small, diffuse; dorsum with verrucae in three longitudinal bands, two lateral close to parapodial lobes, one mid-dorsal with larger verrucae ( Figure 10A). Parapodial lobes with smaller, fewer verrucae, especially on dorsal cirri. Ventral cirri and ventral surface smooth.
Prostomium fused with peristomium. Palps divergent, rounded, palpostyles small rounded (cirriform in paratype). Lateral antennae placed about the middle of the prostomium, rounded, do not reach anterior margin of palps. Tentacular cirri cirriform, dorsal slightly longer and wider than ventral one, tips bent or slightly eroded.
Parapodial lobes short, as long as one-fourth to one-fifth of body width. First dorsal cirri slightly longer than dorsal cirri of setiger 2. Dorsal and ventral cirri fusiform; dorsal cirri one-tenth longer than ventral cirri. Cirrophore not distinct. No glandular material nor pigment spots. Paratype with ventral cirri longer than setal lobe ( Figure 10B-E). Dorsal cirrophore with slightly smaller verrucae than those present on the back. Coelom with small eggs. Neurosetae few, very finely spinulose unidentate limbates, and many long smooth capillaries.
Posterior end regenerating about eight setigers; pygidium with verrucae and lateral anal cirri, slightly longer than adjacent verrucae (arranged in three or four concentric rows in paratype). Pharynx completely everted; it has a transparent outer layer with a muscular and glandular core that separate in distal and basal portions. Paratype with gut diverticula present from setiger 4; with small eggs in median and posterior setigers, each about 50 mm.

Remarks
Pilargis mohri differs from other species in the genus because it has a definite pattern of verrucae abundance, and parapodia are small, in relation to body width, making it resemble other pilargids with slender bodies, like Loandalia Monro.

Distribution
Restricted to the type locality, Southern Vietnam, Western Pacific Ocean.

Description (translated and modified from Buzhinskaja and Britayev 1994)
Holotype of P. v. pacifica drab yellow, densely verrucose, including prostomium, palps, and cirri; it is 82 mm long, 2.5 mm wide including parapodia, ca. 350 segments. Prostomium divided in three parts: two lateral lobes and one small triangular lobe. Small antennae are attached to the distal ends of the paired lobes. Palps fused to each other, separated medially for less than half their length; palpophore massive, palpostyle small. Lateral pigmented spots visible in prostomium. Tentacular segment larger than following segments, ciliated areas on its dorsal posterior part. Tentacular cirri similar in length, ventral ones slightly thinner. First pair of dorsal cirri longer than the rest.
Parapodia with dorsal side covered by large glandular areas, each with small dark pigment cells. Dorsal cirri about twice as wide as cirrostyle. All setae bidentate, shaft with row of teeth; bidentate tips in longer setae only seen under high magnification. Anal cirri lost. Living worms from Vostok Bay with pink spotty body; mature male greenish with parapodial bases greenish yellow spots. Preserved specimens with small brown spots especially dorsally, prostomium, palps, and tentacular segment spotty. One specimen from Vostok Bay with two anal cirri.

Remarks
With specimens from Northern Japan, Zachs (1933) introduced the new name as a subspecies. He gave as its distinguishing feature the lack of furcate setae that had been repeatedly included for the stem species (Fauvel 1920(Fauvel , 1923(Fauvel , 1934(Fauvel , 1936. As stated above, these setae are not present in the genus, so their differences have to be revised. Further, Pilargis verrucosa pacifica Zachs, 1933, is a nomen nudum after ICZN Art. 13 (ICZN 1999), and thus unavailable. The use by Uschakov in 1955(Uschakov 1965 implies that it should be cited as Pilargis verrucosa pacifica Uschakov, 1955 (ICZN Art. 50.1). However, he failed to indicate the re-establishment of the name, or if he had used the same type materials.
Pilargis pacifica Uschakov, 1955 n. stat. differs from P. verrucosa by having foliose tentacular and dorsal cirri and smaller verrucae over the body, while P. verrucosa has blunt tentacular and dorsal cirri, and larger verrucae over the body. These differences are regarded as sufficient to change its rank. Buzhinskaja (1980, p 43) recorded P. berkeleyae for the Japan Sea; she overlooked the previous papers (Zachs 1933;Uschakov 1965), which were using P. v. pacifica. Later, Buzhinskaja and Britayev (1994, p 95) found in the Vostok Bay what they regarded as the same species; they redescribed the type material of P. v. pacifica available in St Petersburg, and concluded that it was a junior synonym of P. berkeleyae. They regarded Zachs' name as a junior synonym because his paper was published after Monro's and because they found no differences after comparing it with the redescription by Hartman (1947). By comparing those illustrations, however, there are several differences between these two species: P. pacifica has a large tentacular segment, as wide as setiger 1, while in P. berkeleyae it is narrower than setiger 1; the dorsal cirri in P. pacifica have cirrostyles slightly thinner than cirrophores, while they are markedly thinner in P. berkeleyae; and in P. pacifica setal lobes are elongate triangular, while they are short truncate in P. berkeleyae. Therefore, they cannot be synonyms.
In fact, after the available morphological and ecological information available (Imajima 1987;Britayev 1993), there may be more than one species in the Northwestern Pacific. The relationships among these species have to be based on the revision of materials which were not available to us.

Redescription
Holotype (ZMUB-53527) complete, twisted over itself, making difficult any measurement without further damaging it. It was described as 20 mm long (it is 5.7 mm long), 1.0 mm wide (setiger 15), 80 setigers and a regenerating posterior end with eight or nine asetigers; it has 26 setigers and a regenerating portion with two inmature setigers and three preanal asetigers. It is now colorless but the posterior brain lobes are dark (occupy first setiger and slightly invade the second one). Body flat, densely covered by verrucae dorsally, verrucae abundant over tentacular and dorsal cirri ( Figure 11A). Ventrally verrucae restricted to parapodial bases, smaller, leaving a smooth midventral wide area limited by two longitudinal muscle bands. Left setigers 7-10 are folded dorsally (because of body compression in the vial).
Posterior end with pygidium in regeneration; damaged ( Figure 11D); two setigers and three preanal asetigers. Pygidium as an inverted truncate cone, two ventrolateral anal cirri well developed, with many verrucae. Paratypes with enteric diverticula dark, clearly seen from setiger 15; few eggs can be seen from about setiger 50. Posterior fragment shows them too.

Discussion
The holotype differs in several regards to the original description, besides the difference in size. The original illustration included a ventral view, which was indicated as a dorsal view. The posterior brain lobes are dark, there are two other lateral smaller dark glands projecting towards the posterolateral corners of the tentacular segment, and the pigmented glandular area in posterior cirrophores has faded slightly. The dorsal cirri are larger in median setigers; the second dorsal cirri is not shorter than the first dorsal cirri. The neurosetae are distally entire, not bidentate, and have a thin, smooth blade.
This species belongs to the group with abundant dispersed verrucae over the back and parapodial lobes. Pilargis papillata is closely allied to P. modesta and P. rozbaczyloi n. sp., but differs from them by having the first dorsal cirri smaller than the following ones rather than larger, and fusiform verrucose dorsal cirri, contrasting with digitate verrucose cirri in P. rozbaczyloi n. sp., and smooth fusiform cirri in P. modesta.

Distribution
Restricted to the type locality in Southwestern Norway in depths of over 400 m.
First setiger with dorsal cirri as long as dorsal tentacular cirri, longer than dorsal cirri of setiger 2. Anterior setigers with dorsal and ventral cirri elongated, cirrostyles about as long as ventral cirri ( Figure 13A); glands scattered. Median parapodia with notocirri thicker than neurocirri, of about the same length; cirrostyles shorter than in anterior setigers ( Figure 13B), verrucae over dorsal surface of parapodial lobe ( Figure 13E); glands become more abundant and closely packed. Posterior parapodia with dorsal cirrostyles 1.5 times as long as ventral cirri; glands few, some ventral ones aligned along the base of parapodia ( Figure 13D).
Parapodial glands well developed, a rounded basal group over the anterior face of dorsal cirrophore, and two other glands placed over ventral surface; present in setigers 9-98 but better developed by setiger 50. Glands in median segments with pigment diffuse, better developed in notopodial base, and ventrally along parapodial base; additional glands smaller, diffuse ( Figure 13C). Neurosetae of two types, capillaries long or short, limbates denticulated, distally curved, bidentate ( Figure 13F).
Posterior end abruptly tapered giving the appearance of being regenerated; pygidial bulb well developed, as long as previous four asetigers, covered with coarse verrucae over the distal margin and with a proximal, slender almost smooth surface. Two ventrolateral anal cirri covered by truncated verrucae ( Figure 12B). Two paratypes had different pygidia; one had a non-expanded anal bulb with two verrucate anal cirri ( Figure 12C), while another one lacked anal cirri and distal expansion, but the verrucae had a similar abundance ( Figure 12D).

Discussion
Pilargis rozbaczyloi n. sp. resembles P. berkeleyae; they differ because in P. berkeleyae verrucae are generally small, restricted to the anterior end and parapodial lobes, the cirrophore is much larger than the cirrostyles, and the glandular region on cirrophores is massive. Cañ ete et al . (1990) material was not available; their illustrations and description are good enough to be confident about the identity of their materials. The development of gland areas in P. rozbaczyloi n. sp. resembles P. maculata, being restricted to the anterior side of the cirrophore, but verrucae are larger in the former, and both differ from P. berkeleyae because its glands tend to cover the whole cirrophore. The single specimen found by Dean (1999) might belong to this species.

Etymology
The species is named after Nicolás Rozbaczylo because of his important work and publications on Chilean polychaetes, and for helping to make this material available.

Distribution
Northern and Central Chile in 25-70 m water depth.

Description
Body very long, with 320 setigers. Prostomium separated from peristomium; palps biarticulated, directed ventrally, palpostyles small digitate, slightly larger than verrucae. Tentacular cirri unequal, dorsal one about five times as long as ventral cirri. First setiger with dorsal cirri twice as long as those of setiger 2. All setigers with globose dorsal cirri, tapering, larger than ventral cirri; glands small on anterior parapodia. Size of glands in median and posterior setigers similar (after the original drawings). Verrucae apparently restricted to dorsal cirri.

Remarks
The original illustration shows very large dorsal tentacular cirri and dorsal cirri of setiger 1. In fact, this has been used to separate the species from the others (Pettibone 1966). Pilargis tardigrada is closely allied to P. maculata; they differ because in P. tardigrada dorsal cirrostyles are fusiform and notopodial glands are dispersed, while in P. maculata dorsal cirrostyles are digitiform and glands are concentrated over the anterior notopodial face.

Description
Holotype complete, pale, twisted, dark glands placed basally on anterior surface of dorsal cirrophore, few ventral glands on parapodial ventral side. Verrucae coarse, over all dorsal surface. It is 17 mm long, 0.8 mm wide, 104 setigers, four asetigers.
First setiger with dorsal cirri about as long as dorsal cirri of setiger 2. All setigers with globose dorsal cirri, acuminate, larger than ventral cirri; glands small on anterior parapodia ( Figure 14B), become larger in median ones ( Figure 14E), less developed in posterior setigers ( Figure 14F). Verrucae present over the parapodia and on ventral surface, larger by median setigers, reduce its size towards posterior end.
Parapodial glands dark, well developed in an elliptic basal group over anterior surface of dorsal cirrophores, two other smaller ones over ventral surface. Start by setiger 5, become larger and more abundant by setiger 10. By setiger 20, become twice as big as anterior glands; by setiger 50, fade out almost completely ( Figure 14C). Neurosetae long or short capillaries and limbates, limbus denticulated, distally curved, entire.

Discussion
Pilargis wolfi n. sp. resembles P. modesta by having non-conspicuous dorsal verrucae but it differs by having well-developed verrucae over median and posterior parapodia. It has a pigmented glandular pattern similar to that found in P. maculata and P. rozbaczyloi n. sp.; it differs from the former by having longer cirrostyles and larger verrucae in median and posterior segments, while from the latter it differs by having a short cirrophore and especially by lacking anterior pigmented glands over posterior setigers.

Etymology
This species is named after Dr. Paul S. Wolf, who was an important collaborator on the Taxonomic Guide to the Polychaetes of the Northern Gulf of Mexico, and because of his important publications on pilargid polychaetes.

Type locality
Off Texas, USA.

Discussion
This fragment does not fit into P. papillata Rasmussen, which has been described from a fjord environment, and it might be an undescribed species. It differs by having many verrucae with an 8-shaped profile, that are abundant and large over cirrostyles, and by having all setae capillaries. More and better material is required to describe the species.
Key to species of Pilargis de Saint-Joseph, 1899